Prev: Optimizer 10053 trace file, strange choises made by Oracle??
Next: Compression
From: Thomas Kellerer on 13 Dec 2009 15:56
Kenneth Koenraadt wrote on 13.12.2009 21:14:
> That's interesting : The online doc clearly states that UPDATE works
> just as INSERT's regarding error logging..however, I have this simple
> test case, which works with INSERT's, but not UPDATE's. The OP wanted
> just that, so it may not work for him. Any thoughts ?

That's in fact documented:

http://download-uk.oracle.com/docs/cd/B19306_01/server.102/b14200/statements_9014.htm#BGBEIACB

"The following conditions cause the statement to fail and roll back without invoking the error logging capability:"
[...]
- Any update operation UPDATE or MERGE that raises a unique constraint or index violation).

I didn't see that either the first time...

Thomas
From: Steve Howard on 13 Dec 2009 19:21
On Dec 12, 1:41 am, Mladen Gogala <gogala.mla...(a)gmail.com> wrote:
>
> Why not simply use the merge statement?
>
> --http://mgogala.byethost5.com

This is a regular event in our shop. Our developers used to try to
insert, and then handle the exception if it existed and update it. We
cut our redo by a large factor when we asked them to switch to MERGE.
From: jimmyb on 14 Dec 2009 00:53
On Dec 13, 4:21 pm, Steve Howard <stevedhow...(a)gmail.com> wrote:
> On Dec 12, 1:41 am, Mladen Gogala <gogala.mla...(a)gmail.com> wrote:
>
>
>
> > Why not simply use the merge statement?
>
> > --http://mgogala.byethost5.com
>
> This is a regular event in our shop.  Our developers used to try to
> insert, and then handle the exception if it existed and update it.  We
> cut our redo by a large factor when we asked them to switch to MERGE.

They used to do an UPDATE when an exception was raised? That should be
a bad coding practice in every shop, JMO.
From: Mladen Gogala on 14 Dec 2009 12:45
On Sun, 13 Dec 2009 21:53:54 -0800, jimmyb wrote:


> They used to do an UPDATE when an exception was raised? That should be a
> bad coding practice in every shop, JMO.

Why is that?



--
http://mgogala.byethost5.com
From: Shakespeare on 14 Dec 2009 13:39
jimmyb schreef:
> On Dec 13, 4:21 pm, Steve Howard <stevedhow...(a)gmail.com> wrote:
>> On Dec 12, 1:41 am, Mladen Gogala <gogala.mla...(a)gmail.com> wrote:
>>
>>
>>
>>> Why not simply use the merge statement?
>>> --http://mgogala.byethost5.com
>> This is a regular event in our shop. Our developers used to try to
>> insert, and then handle the exception if it existed and update it. We
>> cut our redo by a large factor when we asked them to switch to MERGE.
>
> They used to do an UPDATE when an exception was raised? That should be
> a bad coding practice in every shop, JMO.

Not if the exception is a dup-val-on-index. Its a practice seen all around!

Shakespeare
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4
Prev: Optimizer 10053 trace file, strange choises made by Oracle??
Next: Compression